cnn.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "A100",
      "authorship_tag": "ABX9TyPBytgi8tSRKEHvhkQ0gQxB",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/gandhiUH/textProcessing/blob/master/cnn.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lOIEGsitwWkb",
        "outputId": "15dc6019-364e-4625-be34-4cf52cfeaf8c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Found GPU at: /device:GPU:0\n"
          ]
        }
      ],
      "source": [
        "import tensorflow as tf\n",
        "device_name = tf.test.gpu_device_name()\n",
        "if device_name != '/device:GPU:0':\n",
        "  raise SystemError('GPU device not found')\n",
        "print('Found GPU at: {}'.format(device_name))"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "from __future__ import print_function\n",
        "import keras\n",
        "from keras.models import Sequential\n",
        "from keras.utils import to_categorical\n",
        "from keras.layers import Dense, Conv2D, MaxPooling2D, Dropout, Flatten"
      ],
      "metadata": {
        "id": "2x4bg0Lewziq"
      },
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.datasets import cifar10\n",
        "(train_images, train_labels), (test_images, test_labels) = cifar10.load_data()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vqAzBLjaw6h5",
        "outputId": "241f670c-22d9-4976-cc90-d6c871fa8e60"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n",
            "170498071/170498071 [==============================] - 6s 0us/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print('Training data shape : ', train_images.shape, train_labels.shape)\n",
        "\n",
        "print('Testing data shape : ', test_images.shape, test_labels.shape)\n",
        "\n",
        "# Find the unique numbers from the train labels\n",
        "classes = np.unique(train_labels)\n",
        "nClasses = len(classes)\n",
        "print('Total number of outputs : ', nClasses)\n",
        "print('Output classes : ', classes)\n",
        "\n",
        "plt.figure(figsize=[4,2])\n",
        "\n",
        "# Display the first image in training data\n",
        "plt.subplot(121)\n",
        "plt.imshow(train_images[0,:,:], cmap='autumn')\n",
        "plt.title(\"Ground Truth : {}\".format(train_labels[0]))\n",
        "# Display the first image in testing data\n",
        "plt.subplot(122)\n",
        "plt.imshow(test_images[0,:,:], cmap='gray')\n",
        "plt.title(\"Ground Truth : {}\".format(test_labels[0]))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 312
        },
        "id": "ff618yV-xClJ",
        "outputId": "fc0cea87-fb93-4d70-dcfa-c62015aa34c2"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training data shape :  (50000, 32, 32, 3) (50000, 1)\n",
            "Testing data shape :  (10000, 32, 32, 3) (10000, 1)\n",
            "Total number of outputs :  10\n",
            "Output classes :  [0 1 2 3 4 5 6 7 8 9]\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Ground Truth : [3]')"
            ]
          },
          "metadata": {},
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 400x200 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "nRows,nCols,nDims = train_images.shape[1:]\n",
        "print(train_data.shape)\n",
        "print(train_data.shape[0:])\n",
        "print(train_data.shape[1:])\n",
        "print(\"**********\")\n",
        "print(train_data.shape)\n",
        "print(test_data.shape)\n",
        "\n",
        "train_data = train_images.reshape(train_images.shape[0], nRows, nCols, nDims)\n",
        "test_data = test_images.reshape(test_images.shape[0], nRows, nCols, nDims)\n",
        "input_shape = (nRows, nCols, nDims)\n",
        "\n",
        "train_data = train_data.astype('float32')\n",
        "test_data = test_data.astype('float32')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XNjGR456y9KQ",
        "outputId": "3f29b654-aa87-4599-e55b-aef3a77384cf"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(50000, 32, 32, 3)\n",
            "(50000, 32, 32, 3)\n",
            "(32, 32, 3)\n",
            "**********\n",
            "(50000, 32, 32, 3)\n",
            "(10000, 32, 32, 3)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_data /= 255\n",
        "test_data /= 255\n",
        "\n",
        "train_labels_one_hot = to_categorical(train_labels)\n",
        "test_labels_one_hot = to_categorical(test_labels)\n",
        "print(train_labels_one_hot)\n",
        "print(test_labels_one_hot)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eoQozm6c0Onf",
        "outputId": "e9a6e6bf-815e-4f79-c989-090fb4228b7a"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[[0. 0. 0. ... 0. 0. 0.]\n",
            " [0. 0. 0. ... 0. 0. 1.]\n",
            " [0. 0. 0. ... 0. 0. 1.]\n",
            " ...\n",
            " [0. 0. 0. ... 0. 0. 1.]\n",
            " [0. 1. 0. ... 0. 0. 0.]\n",
            " [0. 1. 0. ... 0. 0. 0.]]\n",
            "[[0. 0. 0. ... 0. 0. 0.]\n",
            " [0. 0. 0. ... 0. 1. 0.]\n",
            " [0. 0. 0. ... 0. 1. 0.]\n",
            " ...\n",
            " [0. 0. 0. ... 0. 0. 0.]\n",
            " [0. 1. 0. ... 0. 0. 0.]\n",
            " [0. 0. 0. ... 1. 0. 0.]]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print('Original label 0 : ', train_labels[0])\n",
        "print('After conversion to categorical ( one-hot ) : ', train_labels_one_hot[0])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ONPxZgIg1kcG",
        "outputId": "ae940e37-786b-47e5-b144-e9fdb0d702d8"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Original label 0 :  [6]\n",
            "After conversion to categorical ( one-hot ) :  [0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def createModel():\n",
        "    model = Sequential()\n",
        "    # The first two layers with 32 filters of window size 3x3\n",
        "    model.add(Conv2D(32, (3, 3), padding='same', activation='relu', input_shape=input_shape))\n",
        "    model.add(Conv2D(32, (3, 3), activation='relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "    model.add(Dropout(0.25))\n",
        "\n",
        "    model.add(Conv2D(64, (3, 3), padding='same', activation='relu'))\n",
        "    model.add(Conv2D(64, (3, 3), activation='relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "    model.add(Dropout(0.25))\n",
        "\n",
        "    model.add(Conv2D(64, (3, 3), padding='same', activation='relu'))\n",
        "    model.add(Conv2D(64, (3, 3), activation='relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "    model.add(Dropout(0.25))\n",
        "\n",
        "    model.add(Flatten())\n",
        "    model.add(Dense(512, activation='relu'))\n",
        "    model.add(Dropout(0.5))\n",
        "    model.add(Dense(nClasses, activation='softmax'))\n",
        "\n",
        "    return model"
      ],
      "metadata": {
        "id": "NucDH2OY15CU"
      },
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model1 = createModel()\n",
        "batch_size = 256\n",
        "epochs = 50\n",
        "model1.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])"
      ],
      "metadata": {
        "id": "F8UY-1oY26rc"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model1.summary()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oQA4s6EL3GG_",
        "outputId": "a770bf6f-3163-4c32-9262-e5f09db0f46e"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " conv2d (Conv2D)             (None, 32, 32, 32)        896       \n",
            "                                                                 \n",
            " conv2d_1 (Conv2D)           (None, 30, 30, 32)        9248      \n",
            "                                                                 \n",
            " max_pooling2d (MaxPooling2D  (None, 15, 15, 32)       0         \n",
            " )                                                               \n",
            "                                                                 \n",
            " dropout (Dropout)           (None, 15, 15, 32)        0         \n",
            "                                                                 \n",
            " conv2d_2 (Conv2D)           (None, 15, 15, 64)        18496     \n",
            "                                                                 \n",
            " conv2d_3 (Conv2D)           (None, 13, 13, 64)        36928     \n",
            "                                                                 \n",
            " max_pooling2d_1 (MaxPooling  (None, 6, 6, 64)         0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " dropout_1 (Dropout)         (None, 6, 6, 64)          0         \n",
            "                                                                 \n",
            " conv2d_4 (Conv2D)           (None, 6, 6, 64)          36928     \n",
            "                                                                 \n",
            " conv2d_5 (Conv2D)           (None, 4, 4, 64)          36928     \n",
            "                                                                 \n",
            " max_pooling2d_2 (MaxPooling  (None, 2, 2, 64)         0         \n",
            " 2D)                                                             \n",
            "                                                                 \n",
            " dropout_2 (Dropout)         (None, 2, 2, 64)          0         \n",
            "                                                                 \n",
            " flatten (Flatten)           (None, 256)               0         \n",
            "                                                                 \n",
            " dense (Dense)               (None, 512)               131584    \n",
            "                                                                 \n",
            " dropout_3 (Dropout)         (None, 512)               0         \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 10)                5130      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 276,138\n",
            "Trainable params: 276,138\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "history = model1.fit(train_data, train_labels_one_hot, batch_size=batch_size, epochs=epochs, verbose=1,\n",
        "                   validation_data=(test_data, test_labels_one_hot))\n",
        "model1.evaluate(test_data, test_labels_one_hot)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kdaGduwM3Tuv",
        "outputId": "8737bcb9-9fe5-4414-de8a-657edacb0231"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "196/196 [==============================] - 14s 12ms/step - loss: 2.3030 - accuracy: 0.0990 - val_loss: 2.3027 - val_accuracy: 0.1000\n",
            "Epoch 2/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3030 - accuracy: 0.0977 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 3/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3028 - accuracy: 0.0987 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 4/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3029 - accuracy: 0.0979 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 5/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0997 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 6/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3028 - accuracy: 0.0960 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 7/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3028 - accuracy: 0.0985 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 8/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0980 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 9/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.1008 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 10/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0959 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 11/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0979 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 12/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.1010 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 13/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0982 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 14/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0971 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 15/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0975 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 16/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0988 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 17/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0972 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 18/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0970 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 19/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0978 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 20/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0987 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 21/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0970 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 22/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0966 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 23/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0978 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 24/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0979 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 25/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0985 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 26/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.1004 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 27/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0989 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 28/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0976 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 29/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0995 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 30/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0979 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 31/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0989 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 32/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0977 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 33/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0990 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 34/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0974 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 35/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0991 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 36/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0981 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 37/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0971 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 38/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0985 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 39/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0985 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 40/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0988 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 41/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0989 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 42/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0976 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 43/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.1000 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 44/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0971 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 45/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0973 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 46/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0975 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 47/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0991 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 48/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0996 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 49/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.1002 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "Epoch 50/50\n",
            "196/196 [==============================] - 1s 7ms/step - loss: 2.3027 - accuracy: 0.0972 - val_loss: 2.3026 - val_accuracy: 0.1000\n",
            "313/313 [==============================] - 1s 2ms/step - loss: 2.3026 - accuracy: 0.1000\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[2.302589178085327, 0.10000000149011612]"
            ]
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    }
  ]
}